Sex differences and the role of estrogens in the immunological underpinnings of Alzheimer's disease

Abstract

Alzheimer's disease (AD) affects women more frequently and more severely than men, but the biological mechanisms underlying these sex differences remain poorly understood. This review integrates recent findings from neuroscience, immunology, endocrinology, and genetics to explore how sex steroid hormones, particularly estrogen, shape neuroimmune responses and influence AD risk. We highlight the pivotal roles of microglia and astrocytes, whose inflammatory and neuroprotective actions are modulated by hormonal fluctuations across the female lifespan, including pregnancy, menopause, and menopausal hormone replacement therapy. Key genetic risk factors, such as apolipoprotein E ε4, show sex‐specific effects on glial activation, tau pathology, and cognitive decline. Furthermore, life‐stage transitions, especially menopause, intersect with changes in brain metabolism, immune signaling, and epigenetic regulation, increasing susceptibility to neurodegeneration in women. We propose a framework for sex‐aware, personalized approaches to AD prevention and treatment. By integrating hormone–immune interactions with genetic and glial biology, this review emphasizes the critical need for sex‐specific models in AD research.

Highlights

• Women develop greater tauopathy, with more cognitive and clinical consequences in Alzheimer's disease (AD).

• Glial activation is adapted by estrogens to shape vulnerability or resilience to AD.

• Sex differences in innate and adaptive immunity could contribute to AD progression.

• Effects of menopausal hormone therapy on immunity in AD remain understudied.

• Future studies to explore sex differences in immune function during AD are needed.

Conceptual framework for sex‐specific Alzheimer's disease (AD) risk. Sex differences in AD vulnerability arise from the intersection of multiple biological systems. Fluctuations in sex hormones across reproductive life stages (e.g., pregnancy, menopause) shape long‐term neuroendocrine tone. These hormonal shifts influence immune modulation, including both peripheral and central immune activity, particularly in glial cells such as microglia and astrocytes. Concurrently, genetic architecture, including sex‐interacting variants such as apolipoprotein E ε4, modifies susceptibility to AD pathology. The dynamic interplay among these systems contributes to a sex‐specific trajectory of AD risk, with distinct implications for disease onset, progression, and therapeutic response in women. (Figure created with BioRender.)

1. INTRODUCTION

Alzheimer's disease (AD) is a progressive neurodegenerative disease that robs individuals of memory, identity, and ultimately, independence. Currently, an estimated 7.2 million Americans live with AD, ¹ and globally, > 55 million people are affected. Strikingly, more than half of those diagnosed with AD are women. ¹ , ² While aging remains the greatest risk factor, this disparity suggests the influence of biological sex‐specific mechanisms that have yet to be fully elucidated.

Traditionally, research has centered on hallmark proteinopathies (amyloid beta [Aβ] plaques and tau neurofibrillary tangles) as defining features of AD pathology. However, accumulating evidence points to the immune system as a central player in disease pathogenesis. Glial cells such as microglia and astrocytes, both essential for maintaining homeostasis and responding to injury, are intimately associated with AD pathology. Moreover, genetic studies have highlighted risk loci related to immune regulation, and several of these show sex‐specific effects. ³ , ⁴ , ⁵ Importantly, immune function shows sex‐specific variation. Women and men differ not only in innate and adaptive immune responses ⁶ but also in the hormonal regulation of these systems across the lifespan. Sex steroids, including estrogens, progesterone, and testosterone, modulate neuroimmune signaling at the cellular and molecular levels, ⁷ , ⁸ , ⁹ and their fluctuations during pregnancy, menopause, and hormone therapy correspond with altered neuroinflammatory profiles. ¹⁰ , ¹¹ , ¹²

This review posits that interactions between sex steroid hormones and immune function contribute significantly to the sex differences observed in AD risk and progression. Unfortunately, existing literature often conflates sex and gender or lacks clear operational definitions, further complicating the field. While we recognize the important role of gender as a social construct, this review focuses specifically on biological sex, operationalized through hormonal profiles and gonadal status. We will use the terms “female” or “male” for data from animal models and “woman” or “man” for human data to refer to biological sex throughout this document.

Importantly, we acknowledge that biological sex is only one axis along which AD risk is stratified. Socially patterned factors, including health‐care disparities, racial and cultural bias, caregiving burden, and unequal access to education and resources, also intersect with sex to shape vulnerability and outcomes in AD. ¹³ , ¹⁴ , ¹⁵ While these social determinants of health fall outside the scope of this review, their influence remains critical and warrants dedicated exploration in future work.

We begin by reviewing sex differences in AD risk, biomarkers, and genetic architecture, including the effects of apolipoprotein E (APOE) and other loci. We then explore the role of sex hormones in shaping immune function, both in the periphery and within the brain, with a particular focus on glial cell responses. We next examine how systemic immune function, and its modulation by pregnancy, menopause, and menopausal hormone therapy (MHT), might influence long‐term brain health. Throughout this review, we emphasize critical knowledge gaps and propose future directions for life stage‐ and sex‐aware AD research. By integrating insights across genetics, glial biology, systemic immunity, and hormonal regulation, our aim is to develop a coherent framework that explains the intersection of sex‐based biological factors in AD pathogenesis and seeks to inform more personalized approaches to its prevention and treatment.

2. SEX DIFFERENCES IN AD PATHOGENESIS

2.1. Risk and prevalence

It is well established that AD prevalence (i.e., the number of AD cases at a snapshot in time) is higher among women, with nearly two thirds of clinical AD cases being women compared to men ¹ , ² (Table 1). However, the literature is mixed regarding sex differences in incidence rates (i.e., new AD cases diagnosed over a period of time). Most studies do not find consistent sex differences in incidence, ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ but other studies find trends, which are observed mainly in older age groups, such as 85+ years of age. ²⁰ , ²¹ , ²² The Framingham Heart Study identified that the overall lifetime risk for AD or dementia is 2‐fold higher in women >65 years of age compared to men. ²³ , ²⁴ This elevated risk in women reflects a combination of biological and physiological factors, not solely longevity. Although age remains the strongest predictor of AD, evidence points to sex‐specific differences in how the disease develops and progresses. These include differential impacts of hormonal transitions, immune function, and genetic vulnerability, which will be explored in the following sections.

TABLE 1.

Sex differences in AD risk, biomarkers, and genetic drivers.

Males vs. females
Risk		References
Prevalence	♂ < ♀	1 , 2
Lifetime risk	♂ < ♀	23 , 24
Incidence	♂ =? ♀ (Most studies show no sex differences, but some show a higher incidence among females.)	16 , 17 , 18 , 19 , 20 , 21 , 22

Biomarkers
Amyloid	♂?< ♀ (Mixed findings on sex differences in amyloid burden; some studies show a trend for higher burden in females.)	25 , 26 , 27
Tau	♂ < ♀	25 , 26 , 27 , 28 , 29 , 30 , 31 , 32
Response to pathology	♂ < ♀	4 , 5 , 33 , 34 , 35 , 36 , 37 , 38

Genetic drivers of biomarkers
APOE ε4	AD risk: ♂ < ♀ (Mainly between the ages of 65–75)	39 , 40
	Pathology: ♂ < ♀	41 , 42
	Cognition: ♂ < ♀	43
APOE ε2	♂ =? ♀ (Some mixed findings)	44
Beyond APOE	Sex‐specific loci associated with:
	Memory	45
	Cognitive resilience	46
	Neuropathology	36 , 47
	Biomarkers	48

Abbreviations: AD, Alzheimer's disease; APOE, apolipoprotein E.

2.2. Biomarkers and clinical expression

Post mortem studies, biomarker analyses, and neuroimaging data consistently show that women carry a greater burden of AD neuropathology than men, especially with regard to tau accumulation. Multiple investigators report higher densities of tau tangles in women compared to men, ²⁵ , ²⁸ as well as stronger tau signals by positron emission tomography (PET). ²⁹ , ³⁰ , ³¹ This sex difference in tau PET becomes apparent among postmenopausal cognitively normal women versus men. ³⁰ , ³¹ MHT reduces this elevated tau PET signal in women, but the reduction is modest, and men still exhibit lower tau PET signal compared to women. ³⁰ Evidence for sex differences in Aβ is more mixed. While some studies do show subtle differences in Aβ burden between men and women, ²⁵ , ²⁶ others suggest Aβ may be more important in modulating sex differences in tau. ²⁶ For example, there is a stronger association between PET amyloid positivity and PET tau burden in women compared to men. ²⁶ Furthermore, amyloid‐positive women show higher entorhinal cortical tau than amyloid‐positive men, ²⁶ and show a faster rate of tau accumulation when Aβ is present. ²⁶ , ³⁸ It has also been proposed that menopause may influence Aβ levels, with sex differences emerging during peri‐ and postmenopausal stages. ²⁷

Few studies have examined sex differences in biomarkers specifically related to immune activation in AD. PET tracers recognizing the mitochondrial translocator protein (TSPO) are used as markers of neuroinflammation and glial activation in AD. ⁴⁴ , ⁴⁹ In cognitively normal individuals, women show higher total volume of distribution for a TSPO ligand (TSPO signal) than men. ⁵⁰ Among individuals with AD, women demonstrate higher cortical TSPO PET z scores versus men, ⁵¹ and microglial activation is more closely linked with tau PET signal in women. ³⁷ In addition, Bernier et al. ³ describe a sex‐by‐soluble tumor necrosis factor receptor (sTNFR2) interaction in cerebrospinal fluid (CSF), such that higher sTNFR2 levels are associated with worse cognition in women only, an effect that may be mediated by tau levels. ³

Sex differences also appear in the downstream cognitive and clinical consequences of AD pathology. AD biomarker positive women experience more rapid hippocampal atrophy and cognitive decline, ⁴ , ⁵ , ³⁴ , ³⁵ , ³⁷ and the presence of AD biomarkers is more likely to lead to a diagnosis of clinical dementia in women than in men. ³⁷ For a given level of neuropathology, women have a > 20‐fold increased risk for clinical AD compared to a 3‐fold risk in men. ³⁶ , ³⁷ , ⁵² Moreover, cognition at final clinical assessment prior to death correlates more strongly with neuropathological burden at autopsy in women than in men. ³⁷ Together, these findings demonstrate clear sex differences in AD biomarkers. Tau appears to be the most consistent differentiator between women and men in terms of both accumulation and clinical impact. These patterns underscore the importance of prioritizing sex as a biological variable in biomarker research, particularly in relation to tau dynamics and their clinical consequences.

2.3. Genetic modulators of AD risk and biomarkers

2.3.1. Sex differences in APOE, a known AD risk and immune‐related locus

Recent evidence underscores that sex‐specific genetic drivers may contribute to sex differences in the AD neuropathological cascade. First, polymorphism at the APOE locus is the most robust genetic risk factor for late‐onset AD. In humans, APOE exists in three common allelic forms, ε2, ε3, and ε4, each differing by a single amino acid and conferring different levels of AD risk. The ε3 allele is the most prevalent in the general population and is considered the common variant. The ε4 allele promotes greater risk for AD, while the ε2 allele harbors protective effects. ³⁹ , ⁴⁰ ApoE biology is closely tied to neuroimmune function ⁴⁹ , ⁵³ , ⁵⁴ and its effects on AD risk and pathology show sex‐specific variability. ³⁹ , ⁴⁰ , ⁴¹ , ⁴³ , ⁵⁵ , ⁵⁶ , ⁵⁷ Farrer et al. ³⁹ report a significant sex by APOE genotype interaction, indicating that the relationship between APOE and AD risk differs by sex. Additional work shows that sex differences in the impact of APOE ε4 and AD risk emerge most clearly between the ages of 65 and 75, with women ε4 carriers showing higher AD risk relative to men carrying an ε4 allele. ⁴⁰

Sex differences also appear in the associations of APOE ε4 with clinical AD and AD biomarkers. The presence of a single APOE ε4 allele, observed in 40% to 65% of individuals with AD, ¹ is associated with accelerated Aβ accumulation in women. ³¹ , ³² , ³³ , ⁵⁸ Other studies, however, find no sex difference between APOE ε4 and amyloid burden, suggesting that sex‐related effects of APOE may emerge downstream of amyloidosis. ⁴ , ⁵ , ³⁴ , ³⁵ Hohman et al. ⁴² report that APOE ε4 was more strongly associated with CSF tau in women (across the AD clinical spectrum) compared to men, especially among amyloid‐positive women. Another study reported an APOE‐by‐sex interaction on brain metabolism and cortical thickness, with older APOE ε4 carrier women exhibiting greater brain hypometabolism and cortical thinning compared to APOE ε2 carriers, but the same relationship was not present in men. ⁴¹ Furthermore, a large meta‐analysis found that APOE ε4 was more strongly associated with memory and language performance in women than in men across diverse ancestries. ⁴³

The ε2 allele may also confer sex‐dependent associations with neuroprotection. For instance, Neu et al. ⁴⁰ report that the APOE ε2/ε3 genotype produces stronger protection against AD among women than men. A second cross‐sectional study found that the ε2 allele is more protective against cognitive decline when carried by cognitively normal older women than by men. ⁵⁵ Notably, ε2 carrier status is associated with lower circulating levels of the inflammation marker C‐reactive protein (CRP) in women but not in men. ⁵⁵ However, other work shows the reverse: stronger protective effects of the ε2 allele on cognitive decline in men than in women. ⁵⁶ Finally, three‐way intersectional effects among sex, APOE ε2, and executive function are also observed, such that the ε2 allele exerts a female‐specific protective effect on baseline executive function among cognitively normal older adults, ⁴³ further supporting the complexity of these relationships. Taken together, both ε4 and ε2 alleles exhibit sex‐specific effects on AD risk, biomarker expression, cognition, and inflammation. However, ε2 is relatively rare in the population, which limits statistical power, and future studies are needed to replicate and extend these findings.

2.3.2. Beyond APOE: Other genetic drivers associated with neuroinflammation

Genetic risk for AD extends beyond the APOE locus, and recent evidence illustrates loci that show sex‐specific associations with biomarkers, particularly those related to immune and inflammatory functions. Deming et al. ⁴⁸ re‐analyzed genome‐wide association study (GWAS) data to identify several loci with novel sex‐specific associations. One such association in women only is between amyloidosis and members of the SERPINB gene family, including SERPINB1 and SERPINB6. SERPINB1 has downstream effects on neutrophil accumulation, ⁴⁸ making it a particularly interesting candidate for sex‐specific immune modulation in AD. Other genetic variants are linked to sex‐specific differences in neurofibrillary tau tangles, ³⁶ , ⁴⁷ baseline memory performance, ⁴⁵ and cognitive resilience, ⁴⁶ the latter defined as better‐than‐expected cognition given amyloid burden. Notably, none of these sex‐specific loci were identified in the sex‐agnostic versions of these genetic studies. ⁵⁹ , ⁶⁰ One sex‐specific genetic study found evidence for shared genetic architecture between cognitive resilience and multiple immune traits, including susceptibility to multiple sclerosis, ⁴⁶ suggesting overlapping immune pathways that may differ by sex. Future studies should replicate these findings and continue to investigate the shared biological pathways underlying them.

Overall, within APOE and beyond APOE, sex‐specific genetic factors contribute to sex differences in AD risk, biomarker expression, and cognitive outcomes, supporting the need for sex‐aware genetic studies to better understand disease mechanisms and inform the development of targeted interventions. In the following sections, we will discuss key differences in immune pathways in male and female glial cells.

3. SEX STEROIDS AND THE BRAIN IMMUNE ENVIRONMENT

Sex differences in brain structure, connectivity, and glial cell function are well documented. Perhaps the most obvious biological driver of these differences is the differential expression of the gonadal steroid hormones estradiol (E₂; a form of estrogen), progesterone, and testosterone. While these hormones are synthesized primarily in the ovaries and testes, their small molecular size and lipophilic properties allow them to penetrate the blood–brain barrier (BBB) as well as the plasma membrane of neurons and glial cells. ⁶¹ In addition, steroid hormones are locally synthesized and interconverted in the central nervous system using the same enzymatic pathways as in the ovaries, making the brain a steroidogenic organ. ⁷ , ⁶¹ , ⁶² , ⁶³ , ⁶⁴ Aromatase, the enzyme responsible for converting testosterone into E₂, is localized primarily in neurons, but astrocytes upregulate expression after injury or ischemia. ⁶⁵ , ⁶⁶

Estrogens are widely distributed throughout the brain and regulate a range of higher order functions, including mood, anxiety, fear, and learning and memory. ⁸ In experimental models, estrogens are potent neuroprotectants: they reduce amyloid precursor protein (APP) processing into Aβ, increase clearance of Aβ, suppress tau hyperphosphorylation, and prevent aberrant attempts by neurons to re‐enter the cell cycle. ⁷ , ⁶⁶ , ⁶⁷ , ⁶⁸ These effects have led to the hypothesis that the increased prevalence and severity of AD in women may be due to the postmenopausal decline in gonadal steroid hormones ¹⁰ , ⁶⁹ (see Section 6.2). Neuroimaging studies support this view, showing that biomarkers of preclinical AD in women emerge during the menopausal transition. ⁷⁰

Although circulating levels of estrogens are comparable in cognitively normal women and women with AD, ⁷¹ brain estrogen levels are lower in women with AD. ⁷² Lower aromatase expression in specific brain regions is reported in AD brains, ⁷³ and single nucleotide polymorphisms (SNP) near the aromatase gene CYP19 are associated with AD. ⁷⁴ Additionally, SNPs in estrogen receptor 1 or 2 genes (ESR1, ESR2 encoding estrogen receptor α [ERα] and β [ERβ], respectively) are linked to cognitive decline in both men and women. ⁷⁵ Methylation changes in GPER1, the gene encoding the G protein‐coupled estrogen receptor (GPR30) are associated with reduced cognition, higher tau tangle density, and elevated global AD pathology scores in women, while no such associations are found for ESR1 or ESR2. ⁷⁶ These findings suggest that alterations in brain‐derived estrogen signaling, whether genetic, epigenetic, or biochemical, may contribute to AD vulnerability in women. Future research should map regional neurosteroidogenic activity in the aging human brain and assess whether these local hormone deficits precede or follow early pathological changes.

4. GLIAL CELLS AND SEX DIFFERENCES IN IMMUNE RESPONSE

4.1. Steroid hormone receptor expression in microglia and astrocytes

Both microglia and astrocytes express steroid hormone receptors and exhibit sex differences in morphology, gene expression, and immune function. However, these glial subtypes differ substantially in their receptor profiles, temporal sensitivity to sex hormones, and developmental programming of sex‐specific phenotypes. Adult mouse brain microglia express ERα and ERβ, but not progesterone receptor (PRα) or androgen receptor. ⁷ , ⁷⁷ ERα density does not differ between sexes in adulthood, ⁷⁸ although Esr2 (encoding ERβ) increases with aging. ⁷ , ⁷⁸ Microglia also express the membrane‐bound GPR30, which binds E₂ with 10‐fold less affinity than classical receptors. ⁷ , ⁷⁹

Microglial hormone signaling occurs through both canonical and non‐canonical mechanisms. Canonical estrogen response elements (EREs) are present in promoter sequences of immune‐related genes, such as toll‐like receptor 9 (TLR9), and deletion of the ERα DNA binding domain blocks CpG‐induced TLR9 expression. ⁸⁰ Signaling through ERα is important for the anti‐inflammatory actions of E₂ in microglia because global deletions of ERα, but not ERβ, block activation in response to lipopolysaccharide (LPS) in mice. ⁸¹ In addition, phosphorylation of ERα at Ser216 is critical for anti‐inflammatory signaling: a microglial‐specific knock‐in mouse lacking this phosphorylation site shows enhanced activation markers and increased expression of the pro‐inflammatory cytokines tumor necrosis factor alpha (TNFα) and interleukins (IL) 1α, IL‐1β, and IL‐6. ⁸² Non‐canonical signaling pathways also operate. Unliganded ERα can be phosphorylated by epidermal (EGF)‐ or insulin growth factor‐activated kinases (mitogen‐activated protein kinase [MAPK], protein kinase A [PKA], p21 rat sarcoma/rat sarcoma/extracellular signal‐regulated kinase [p21Ras/ERK]). ⁷ Second, E₂ can regulate nuclear factor‐κB (NF‐κB) binding via phosphatidylinositol‐3‐kinase (PI3K)‐dependent, non‐genomic signaling to affect cytokine secretion without direct involvement of gene transcription. ⁷ Finally, E₂ may also modulate immune function through regulation of calcium‐dependent signaling pathways such as calcineurin‐mediated dephosphorylation of the nuclear factor of activated T cell (NFAT). ⁸³ , ⁸⁴ By regulating calcineurin activity, estrogen indirectly controls NFAT nuclear translocation and inflammatory gene expression. ⁸⁵

In contrast, astrocytes express a broader range of steroid hormone receptors, including ERα, ERβ, GPR30, Gq‐coupled membrane estrogen receptor (Gq‐mER), nuclear progesterone receptor (PR), and androgen receptors. ⁸⁶ , ⁸⁷ , ⁸⁸ , ⁸⁹ These receptors regulate gene transcription, structural plasticity, proliferation, and responses to immune challenge. One critical downstream target is glial fibrillary acidic protein (GFAP), an intermediate filament protein that indicates astrocyte activation and is dynamically regulated by E₂ and progesterone. GFAP expression peaks during high estrogen phases of the estrous cycle in female rodents, especially in the hippocampus and hypothalamic arcuate nucleus, and decreases after ovariectomy, a pattern reversed by estrogen or estrogen + progesterone replacement. ⁹⁰ , ⁹¹ , ⁹² , ⁹³ , ⁹⁴ , ⁹⁵ , ⁹⁶ , ⁹⁷ Future studies should explore cell‐specific receptor expression in human brain tissue across life stages, and whether selective receptor modulators could be leveraged to fine‐tune glial immune responses without broadly affecting systemic hormone levels.

4.2. Sex‐specific features of microglia and astrocytes

Sex differences in microglial phenotype appear early in development and persist into adulthood, varying by brain region. In rodents, males display higher microglial density during postnatal development in some brain regions, but not all. ⁷ , ⁹⁸ This pattern reverses in adulthood, with females showing greater density in the same regions. ⁷ , ⁹⁸ Moreover, male microglia tend to exhibit a less activated phenotype compared to females, ⁹⁸ yet they show enhanced antigen presentation, increased expression of major histocompatibility complex class I (MHC‐I), and greater responsiveness to purinergic signaling ⁹⁹ (Figure 1). Female microglia, on the other hand, adopt a more activated profile, ⁹⁸ , ⁹⁹ producing more pro‐inflammatory cytokines and chemokines, such as TNFα, IL‐1β, and C‐X‐C motif chemokine ligand 10 (Cxcl10), compared to male microglia. ⁹⁸ , ¹⁰⁰ , ¹⁰¹ These sex‐specific traits emerge earlier in females and are delayed in males unless exposed to an immune stimulus like LPS, which accelerates maturation in males, but not in females. ¹⁰² Not all microglial features are sex dependent, however. For example, phagocytosis and process complexity appear largely invariant across sexes under homeostatic conditions. ⁹⁹ , ¹⁰³ Surprisingly, these features remain to be fully characterized in human samples.

FIGURE 1.

Synopsis of major sex differences in microglia, astrocytes, and cells of the adaptive immune system. As discussed in the text, there are major differences in the phenotype of brain glial cells depending on sex, which produces distinctive functional outcomes. All entries refer to rodent literature unless otherwise indicated. Figure generated with BioRender.com. Cxcl10, C‐X‐C motif chemokine ligand 10; E2, estradiol; IL, interleukin; NF‐κB, nuclear factor‐κB; NK, natural killer; TLR, toll‐like receptor; TNFα, tumor necrosis factor alpha.

Importantly, many sex differences in microglia persist independently of hormone levels: female‐derived microglia maintain their transcriptional profiles even when transplanted into male brains, and short‐term E₂ replacement does not reverse these signatures. ⁷ , ¹⁰⁴ In contrast, astrocytic sex differences are closely tied to circulating hormones and are more dynamic across the lifespan. E₂ and progesterone increase astrocyte proliferation, upregulate connexin‐43 expression (important for gap junction coupling), and downregulate ERK1/2, particularly in female astrocytes. ¹⁰⁵ , ¹⁰⁶ , ¹⁰⁷ Astrocyte resistance to injury is also sex specific; female astrocytes display reduced cell death after oxygen‐glucose deprivation, an effect abolished in aromatase knockout models. ¹⁰⁸ Astrocytes from males or androgenized females produce higher levels of IL‐1β, IL‐6, and TNFα in response to LPS, ¹⁰⁹ and display elevated calcium currents after ischemic insult. ¹¹⁰

These distinctions underscore fundamentally different, yet complementary, mechanisms by which glial cells integrate sex and hormonal cues. Microglia are developmentally programmed for sex‐specific immune reactivity, with their phenotypes maintained through intrinsic transcriptional and epigenetic mechanisms that persist even in the absence of circulating hormones. Astrocytes, on the other hand, are highly responsive to hormonal fluctuations across the lifespan and display rapid, region‐specific remodeling in response to changes in estrogen, progesterone, and androgens. Together, these glial populations shape the sex‐specific neuroimmune environment in ways that may predispose females to heightened inflammatory reactivity during key life stages, particularly the menopausal transition. Understanding how these glial‐specific, hormone‐sensitive pathways interact over time and how they intersect with genetic and metabolic risk factors like APOE will be critical for identifying new targets for sex‐tailored interventions in aging and AD. These foundational differences in glial biology set the stage for more complex, life stage–specific modulations that will be discussed in Section 6.

4.3. Sex‐specific gene expression changes in microglia and astrocytes

Studies across rodent models and human brain tissue reveal extensive sex differences in gene expression in both microglia and astrocytes, particularly in the context of aging and neurodegenerative disease. However, the depth and scope of evidence for microglia currently outpaces that for astrocytes. In adult male mice, microglia in the cerebral cortex and hippocampus express higher levels of genes associated with the NF‐κB pathway, leukocyte migration, chemotaxis, and MHC‐I and MHC‐II genes. ⁹⁹ By contrast, female microglia express higher levels of genes linked to cytoskeletal remodeling, vascular development, nervous system development, and cell migration. ⁹⁹ Additional studies show female microglia also displayed elevated expression of genes related to inflammatory responses, apoptosis, and sensitivity to LPS. ¹¹⁶ In aging, female hippocampal microglia adopt more pronounced disease‐associated phenotypes and exhibit stronger senescence‐related transcriptional signatures than males. ¹⁰⁰

Hormonal status also modulates gene expression in both glial types. In female rats, APOE RNA increases during proestrus when estrogen levels peak. ¹¹⁷ In vitro experiments confirmed that this increase occurs in both microglia and astrocytes. ¹¹⁷ In addition, microglia from female APP/PS1 transgenic, amyloid overexpression mice display a more activated morphology and upregulate transcripts like triggering receptor expressed on myeloid cells 2 (Trem2), tyrosine receptor binding protein (Tyrobp), and chemokine ligand 6 (Ccl6), which correlates with greater amyloid accumulation and cognitive impairment. ¹¹⁸ Similar sex biases in microglial activation are observed in 5xFAD ¹¹⁹ and APP^{NL‐G‐F 120} mouse models of amyloid accumulation, although interpretation of 5xFAD findings is complicated by the presence of an ERE in the Thy1 promoter that elevates Aβ production in females. ¹¹⁸ , ¹²¹

Not all findings are consistent, however. Hammond et al. ¹²² report no sex differences in microglial diversity or subpopulation abundance across nine transcriptionally distinct clusters. Human studies show that aged women without cognitive impairment exhibit more proinflammatory microglial gene expression in the entorhinal cortex. ⁷⁸ , ¹⁰¹ Similar signatures are seen in peripheral monocyte‐derived macrophage (MDM) cells and induced pluripotent stem cells (iPSCs)‐derived microglia. ¹⁰¹ Importantly, neither menstrual phase nor E₂ treatment in vitro modified these profiles, further supporting developmental or epigenetic origins of sex‐specific gene expression. ¹⁰¹ Additional contributions likely come from sex‐specific epigenetic markers ⁷⁶ , ¹²³ and miRNA. ¹⁰³

Microglial responses to disease also vary by subtype. Single cell RNA sequencing data reveal numerous phenotypic states of microglia across various mouse models and human brain during aging and AD (reviewed in Boche and Gordon ¹²⁴ ). In health, homeostatic microglia adopt a ramified morphology and express microglial‐specific genes, including purinergic receptor P2Y12 (P2RY12), transmembrane protein 119 (TMEM119), colony stimulating factor‐1 receptor (CSF1R), and C‐X3‐C motif chemokine receptor 1 (CX3CR1, fractalkine receptor), with a relatively small population of cells expressing low levels of AD‐related risk genes. With disease, microglia transition into an activated state variously called disease‐associated microglia (DAM), microglia associated with neurodegeneration (MGnD), or activated response microglia (ARM). ⁵³ , ⁵⁴ , ¹²⁰ , ¹²⁵ Although gene sets identifying these subsets vary among investigators, they share core features, including increased expression of AD risk genes and microglial activation markers concomitant with reductions in homeostatic gene expression. Other microglial subtypes include interferon response (IRM) and proliferating microglia. ¹²⁰ , ¹²⁵ , ¹²⁶ Microglial subtypes responding to amyloid appear somewhat different in humans. In a xenograft model using iPSC‐derived human microglia, Mancuso et al. ¹²⁶ identified homeostatic, DAM, and IRM subtypes, but also additional phenotypes emerge, including human leukocyte antigen (HLA), cytokine response 1, cytokine response 2, transitioning cytokine response, and ribosome response microglia. ¹²⁶ Although research into sex differences among these human microglial subtypes is nascent, recent work shows that female APOE ε4 carriers with AD have elevated numbers of microglia expressing DAM‐like genes. ¹²⁷

In contrast to microglia, astrocyte gene expression in aging and AD has received less attention, and explicit reports of sex‐specific patterns are sparse. In APP/PS1 mice, aging increases expression of astrocytic genes associated with perisynaptic remodeling, synapse elimination, mitochondrial maintenance, PI3K/AKT signaling, and endosomal trafficking. ¹²⁸ , ¹²⁹ , ¹³⁰ Proinflammatory astrocytic transcripts include APOE, C4b, and IL33. ¹²⁸ In post mortem human cortex, reactive astrocytes near plaques exhibit elevated immunoreactivity for GFAP, chitinase‐3‐like protein 1 (CHI3L1/YKL‐40), vimentin, and TSPO, consistent with activation. ¹³¹ , ¹³² , ¹³³ However, the lack of sex‐stratified single‐cell analyses remains a major gap.

The sex‐specific transcriptional landscapes of microglia and astrocytes represent a key intersection of neuroimmune regulation and AD risk. Microglia, particularly in females, appear primed to adopt inflammatory and disease‐associated states with age and pathology. Astrocytes, while less characterized transcriptionally, respond dynamically to hormonal changes and injury, especially through glutamate handling and immune regulation. Moving forward, integrated single‐cell multi‐omics across sexes and life stages will be critical to decipher how these glial populations interact in a sexually dimorphic manner to shape vulnerability or resilience to AD.

4.4. Hormonal modulation of glial inflammation and neuroprotection

Sex steroid hormones, particularly estrogens and progesterone, exert profound effects on glial function, with clear evidence for modulation of immune signaling, morphological changes, and survival under stress. These effects are cell‐type specific and often sex specific, suggesting a mechanistic role in the heightened susceptibility of the female brain to AD. Microglia respond robustly to estrogenic modulation. Short exposures to immune stressors such as LPS, CpG oligodeoxynucleotides, hypoxia, or Aβ trigger microglial activation, yet estrogens mitigate these responses (Figure 2). In rodents, E₂ blocks the LPS‐induced increase in activated microglia in both sexes. ⁷ Synthetic (tamoxifen, raloxifene) and natural (genistein) ER ligands dose dependently attenuate Aβ‐ or LPS‐induced production of proinflammatory cytokines, including TNFα, IL‐1β, and CXCL2 by cultured microglia (reviewed by Villa et al. ⁷ ). Conversely, depletion of circulating estrogens via ovariectomy or disruption of estrogen signaling through genetic deletion of ERα in rodents increases expression of microglial markers of activation (e.g., macrophage antigen complex‐1 [MAC‐1]) and inflammatory genes (e.g., Cd11b, Fcgr2b, and Cd45). ⁷ , ⁷⁸ These effects are exaggerated in aged mice and are paralleled by transcriptomic profiles in post mortem human cortex from postmenopausal women, which show elevated levels of CD11B, FCGR2B, CD45, complement (C1Q, C3) and TLR genes (TLR4, TLR9). ⁷ , ⁷⁸ E₂ replacement reverses these changes, underscoring its regulatory capacity. ⁷ , ⁷⁸ , ¹³⁴

FIGURE 2.

Effects of estradiol (E₂) and ovariectomy (OVX) in rodents on microglia and astrocytes. E2 exerts effects on cells consistent with neuroprotection in model systems, while ovariectomy results in increased markers of neuroinflammation. All entries refer to rodent literature. Figure generated with BioRender.com. Aβ, amyloid beta; GFAP, glial fibrillary acidic protein, an intermediate filament in astrocytes; MAC‐1, macrophage‐1 antigen (also known as complement receptor [CR3] consisting of CD11b (integrin αM) and CD18 (integrin β2).

Despite their anti‐inflammatory effects, estrogens do not uniformly attenuate Aβ pathology. E₂ treatment enhances Aβ uptake in microglia‐like N9 cells, but reduces degradation. ¹³⁴ In APP23 transgenic mice, ovariectomy increases the density of MAC‐1–positive microglia surrounding Aβ deposits, a phenotype reversed by E₂, but overall amyloid deposit burden remains unchanged. ⁷¹ , ¹³⁹ In rats infused with preformed Aβ fibrils, E₂ promotes the formation of more, smaller Aβ deposits and increased insoluble Aβ levels. ¹³⁴ These findings suggest that reduced microglial activation does not necessarily translate simply to beneficial effects on amyloid deposition. Further complicating this picture is APOE genotype, which modulates estrogen's immunoregulatory efficacy. Microglia from APOE ε4/ε4 targeted gene replacement mice secreted more TNFα, IL‐6, and nitrite after LPS and interferon‐gamma (IFNγ) exposure compared to APOE ε3/ε3 microglia. ¹³⁵ Notably, E₂ attenuates cytokine secretion only in APOE ε3/ε3 microglia, indicating a diminished protective effect of E₂ in the presence of APOE ε4/ε4. ¹⁴⁰

Like microglia, astrocytes also exhibit sex‐specific responses to hormonal and immune stimuli. Female rodent astrocytes take up more glutamate (a critical neuroprotective mechanism), through processes independent of circulating hormones, ¹¹² while male astrocytes are more vulnerable to glutamate‐induced proliferation arrest. ¹¹¹ Inflammatory stimuli such as LPS induce greater expression of IL‐1β, IL‐6, and TNFα in male astrocytes and those from androgenized females compared to non‐androgenized females. ¹⁰⁹ Moreover, after ischemic insult, male astrocytes display heightened calcium currents, suggesting increased inflammatory reactivity. ¹¹⁰ Estrogens and progesterone also directly shape astrocyte morphology and viability. In vitro, these hormones promote astrocyte proliferation, enhance connexin‐43 expression, and reduce ERK1/2 levels, particularly in astrocytes derived from female animals. ¹⁰⁵ , ¹⁰⁶ , ¹⁰⁷ They also protect astrocytes from oxygen–glucose deprivation–induced cell death, an effect lost in aromatase‐knockout models lacking brain‐derived estrogens. ¹⁰⁸

Emerging techniques now allow for more refined analyses of astrocytic estrogen signaling. For instance, Spence et al. ¹⁴¹ report that selective ERα activation is neuroprotective in experimental autoimmune encephalomyelitis. This protection is abrogated in mice with astrocyte‐specific conditional deletion of ERα, but not in mice with neuron‐specific conditional deletion. Brann et al. further demonstrate that loss of aromatase in astrocytes, but not neurons, alters estrogenic regulation of gliosis and neuroprotection. ⁶⁶ The use of these more sophisticated tools will be necessary to tease out the complex roles for individual cell types in responses to estrogens. Collectively, these findings reveal that both microglia and astrocytes are key mediators of steroid hormone signaling in the brain. While estrogens broadly dampen glial activation and promote survival under cellular stress, their effects on disease pathology, particularly Aβ accumulation, are complex and genotype dependent. Moreover, little is known about how estrogenic signaling in microglia influences the development of the tau pathology unique to AD. Astrocytes, though less well studied, show complementary and distinct estrogenic responses, especially in relation to glutamate handling and cellular resilience. Future research will require a move beyond bulk analyses to dissect how glial responses to estrogens vary by sex, genotype, and disease stage, using tools such as cell type–specific receptor deletion and single‐cell ‐omics.

5. SEX‐SPECIFIC IMMUNE RESPONSES

5.1. Adaptive immune responses

Sex differences in the immune system emerge during development and persist throughout life, shaping susceptibility to infections, autoimmune diseases, and inflammatory conditions. ¹¹³ In general, for many infections, females appear to be infected more easily, but males display more severe disease symptoms. ¹¹³ , ¹¹⁴ A major contributor to this divergence is that women mount stronger adaptive immune responses than males, ⁶ including higher antibody production after vaccination compared to men, ¹¹⁵ although this contention has been challenged. ¹⁴² , ¹⁴³ Women are also more likely to develop autoimmune diseases. ¹¹⁴ These immune differences may be influenced in part by sex hormones, which regulate the balance of T helper cell subsets and cytokine production. Additionally, many genes involved in immune function are encoded on the X chromosome, including forkhead box P3 (FOXP3), a master regulator of T regulatory cell (Treg) production and function, genes controlling B and T cell interactions as well as immunoglobulin (Ig) class switching. ⁶

Sex differences are also observed in lymphocyte populations. Men have more natural killer cells and CD8+ T lymphocytes than women, whereas women have more CD4+ T cells, which upregulate more antiviral genes and more proinflammatory cytokines after challenge than T cells from men. ⁶ Cellular immune phenotypes also show AD‐related vulnerability. Elevated levels of Th1 and Th17 CD4+ T cells and increased proinflammatory gene expression in monocytes are associated with greater AD risk and cognitive decline in both human and animal models. ¹⁴⁴ , ¹⁴⁵ Anti‐inflammatory Tregs decline in some human studies, ¹⁴⁵ , ¹⁴⁶ but not all. ¹⁴⁴ However, experimental augmentation of Tregs through peripheral transplantation reduces cognitive impairment, microglial activation, and amyloid burden in transgenic mouse models of AD (APP/PS1 and 3xTg mice). ¹⁴⁷ , ¹⁴⁸ However, these studies failed to examine sexual differences explicitly.

5.2. Innate immune responses

The innate immune system also exhibits sex differences. Cells of the innate immune system express higher levels of the pattern recognition receptor TLR4 in men compared to women. In contrast, cells from women express more TLR7 and TLR8 than men, both of which are encoded by the X chromosome and can evade X‐inactivation, resulting in increased receptor expression and activity in women. ⁶ Other innate immune genes also contain EREs, which contribute to differential gene expression by sex. ⁶ These molecular differences contribute to sex‐specific patterns of neuroinflammation in aging and AD. Thus, women display greater age‐related increases in inflammatory genes than men ¹⁴⁹ and, while both men and women exhibit increased inflammatory gene expression in the hippocampus and entorhinal cortex with age, only women show increases across additional brain regions. ¹⁵⁰ These findings indicate that aging women exhibit a broader spatial distribution of neuroinflammatory gene expression in the brain compared to aging men, which could contribute to sex differences in AD progression.

5.3. Systemic inflammation and AD risk

Peripheral immune signaling, commonly referred to as systemic inflammation, is a significant contributor to neurodegenerative disease risk, including AD, vascular contributions to cognitive impairment and dementia (VCID), and Parkinson's disease, and has been reviewed recently. ¹⁵¹ Elevated circulating levels of proinflammatory cytokines and acute phase proteins, such as IL‐6, CRP, and α1‐antichymotrypsin (SERPINA3), are associated with an increased risk of both AD and all‐cause dementia in multiple large cohort studies. ¹⁵² , ¹⁵³ , ¹⁵⁴ However, this association is not universal; other inflammatory markers, including IL‐12 and IFNγ, do not consistently predict dementia risk. ¹⁵⁵

Sex‐specific associations have also been reported. In an analysis by Varma et al., ¹⁵⁶ α2‐macroglobulin (A2M), a peripheral marker of systemic inflammation, was linked to increased risk of incident AD in men, but not in women. Similarly, Walker et al. ¹⁵⁷ conducted a longitudinal study tracking CRP levels over more than two decades and found that men with chronically elevated CRP were significantly more likely to develop amyloid accumulation in later life compared to males who maintained low CRP. In contrast, women's CRP trajectories were not associated with future amyloid status. ¹⁵⁷ These findings suggest that systemic inflammation may confer differential biological risk for AD depending on sex, even when levels of inflammatory proteins are comparable.

Taken together, these findings from adaptive, innate, and systemic immunity support a model in which peripheral immune activity can shape neurodegenerative trajectories via sex‐specific mechanisms. This hypothesis is especially relevant for understanding women's risk for AD, where life stage–specific immune modulation, such as that occurring during pregnancy, may influence long‐term cognitive and neuroimmune outcomes. These sex–immune interactions may help explain differential vulnerability across the lifespan and highlight a need for sex‐aware biomarker development and intervention strategies targeting the brain–periphery immune axis.

6. LIFE STAGE–SPECIFIC IMMUNE AND HORMONAL MODULATION IN WOMEN

6.1. Parity, gravidity, and length of reproductive period

Pregnancy represents a unique, sex‐specific physiological state marked by profound changes in both the endocrine and immune systems, which may carry long‐term implications for brain health and AD risk. During pregnancy, the maternal immune system undergoes a functional transformation to tolerate the semi‐allogenic fetus. This adaptation includes a marked shift toward immune tolerance, particularly during the first and second trimesters, when the risk of fetal rejection is highest. ¹⁵⁸ , ¹⁵⁹ These immunologic changes are linked to fluctuations in sex steroid hormone levels, particularly progesterone and estrogen, as well as the pregnancy‐specific hormone, chorionic gonadotropin, all of which begin to rise early in gestation. While implantation and parturition are associated with transient pro‐inflammatory responses, the majority of pregnancy is characterized by a systemic anti‐inflammatory profile, including prominent Th2 cytokine signaling (IL‐4, IL‐13), an increase in Tregs, and suppression of pro‐inflammatory Th1 responses. ¹⁶⁰ , ¹⁶¹ Progesterone and its downstream mediator, progesterone‐induced blocking factor (PIBF), appear to play a key role in this process by activating the Janus kinase‐1/Signal transducer and activator of transcription‐6 (Jak1/Stat6) signaling pathway through the IL‐4 receptor, further enhancing Treg expansion and immune tolerance. ¹⁶² , ¹⁶³ The shift in immune tone during pregnancy is accompanied by improvements in inflammatory autoimmune conditions, including rheumatoid arthritis and multiple sclerosis, yet also confers heightened vulnerability to some, but not all, infections due to the immunosuppressive environment. ¹⁵⁸ , ¹⁵⁹ , ¹⁶⁴ Of particular relevance to AD, this temporary but robust immune remodeling is now thought to produce longer term effects on maternal immune and neural systems that may influence neurodegenerative disease risk later in life. ¹⁶⁵ , ¹⁶⁶

Among the most important immunologic features of pregnancy is an increase in Tregs, which begin to expand during early gestation and remain elevated above prepregnancy levels well into late pregnancy and even postpartum. ¹⁶¹ , ¹⁶⁷ , ¹⁶⁸ , ¹⁶⁹ Given their role in suppressing pro‐inflammatory effector T cells, persistent Treg elevation may result in prolonged reductions in peripheral and central inflammatory signaling. Accordingly, pregnancy, even if not full term, can have an extended immunosuppressive effect that may be particularly beneficial in the context of age‐related neuroinflammatory processes and AD pathogenesis.

Research examining the association between parity and dementia/AD risk has yielded mixed results, with some studies citing pregnancy as a protective factor, and others finding a link between pregnancy and increased dementia risk. ¹⁷⁰ , ¹⁷¹ , ¹⁷² This is not surprising, given the array of potentially confounding social and environmental factors that can influence whether a woman becomes pregnant and the number of pregnancies. However, recent studies support the idea that pregnancy history (parity) and reproductive period length influence brain aging and dementia risk through combined immune and hormonal mechanisms. For example, Fox et al. ¹¹ report that a greater cumulative number of months spent in the first trimester, when Tregs expansion is most pronounced, is associated with reduced risk of AD. Interestingly, this association remained significant after adjusting for months spent in the third trimester, suggesting that immunologic factors in early pregnancy may be more relevant to AD risk than later gestational estrogen exposure. ¹¹ Neuroimaging studies further support a neuroprotective role of pregnancy. In middle‐aged women, a larger number of pregnancies is associated with greater gray matter volume in temporal, frontal, and precuneus areas. ¹⁶⁶ A large‐scale magnetic resonance imaging study of <12,000 women showed that parous women exhibited less evidence of brain aging compared to nulliparous counterparts, even after accounting for common genetic variation. ¹⁷³ These findings were recently replicated in older women from the Rotterdam Study, again showing larger total brain volume in parous women. ¹⁷⁴

In addition to immune shifts, pregnancy also leads to a surge in circulating hormones, particularly E₂, which modulates bioenergetics and protect against AD‐relevant pathology (see Sections 3 and 4.4). In rodent models, E₂ blocks Aβ‐induced neuroinflammation and cell death by inhibiting of NF‐κB signaling in microglia. ¹² , ¹³⁶ , ¹⁷⁵ , ¹⁷⁶ Estrogen's role in brain aging is further supported by studies showing that women with a shorter reproductive period (i.e., fewer years between menarche and menopause) are at increased risk for late‐life cognitive impairment and dementia, although there are numerous examples of discrepant findings. ¹⁶⁶ , ¹⁷⁷ , ¹⁷⁸ , ¹⁷⁹ Taken together, this body of evidence suggests that both parity and reproductive period length may influence AD risk via converging immune and hormonal pathways that may shape the trajectory of neuroinflammatory aging and resilience to neurodegeneration in women.

6.2. Menopause

Menopause is a uniquely female neuroendocrine transition, typically occurring between the ages of 45 and 55, marked by the gradual loss of ovarian function and the cessation of menses. The menopausal transition occurs in parallel with aging, but is biologically distinct in its hormonal underpinnings. Although commonly conceptualized as a reproductive milestone, the menopausal transition exerts wide‐ranging effects on the brain, driven primarily by fluctuations and eventual decline in circulating estrogen and progesterone levels. These hormonal changes affect multiple estrogen‐regulated systems and are accompanied by a range of neurological symptoms, including insomnia, depression, subjective memory complaints, and cognitive decline, all of which are associated with increased risk of AD. ¹² , ¹⁸⁰ Furthermore, premature menopause is an established risk factor for dementia ¹⁸¹ and is linked to elevated tau PET signals. ¹⁸¹ , ¹⁸² Declining E₂ levels in the menopausal transition lead to reductions in brain glucose metabolism and mitochondrial respiration in regions vulnerable to AD pathology. ¹² , ²⁷ , ¹⁸³ , ¹⁸⁴ , ¹⁸⁵ , ¹⁸⁶ In compensatory response, the brain begins relying on ketone bodies for energy. However, this metabolic shift comes at a cost: ketogenesis involves catabolism of myelin lipids, which could contribute to white matter volume loss and structural brain aging. ²⁷ , ¹⁸³ , ¹⁸⁷ This state of hypometabolism and increased fatty acid use is associated with greater Aβ accumulation and reduced synaptic plasticity in preclinical models of menopause. ¹² , ¹⁸⁸ These effects are especially detrimental in APOE ε4 carriers, who not only demonstrate impaired astrocytic clearance of oligomeric Aβ ¹⁸⁹ but also produce less total ApoE protein relative to APOE ε2 and APOE ε3 carriers. ¹⁹⁰ Anti‐inflammatory properties of ApoE include inhibition of lymphocyte proliferation, suppression of Ig synthesis, and reduced neutrophil activation. ¹⁹¹ , ¹⁹² , ¹⁹³ Thus, postmenopausal women who are APOE ε4 carriers may be at dual risk due to reduced estrogen‐mediated support for glial function and reduced ApoE‐mediated immune regulation. This combination may foster a heightened neuroinflammatory environment, impair protein clearance mechanisms, and accelerate proteotoxic burden and cognitive decline.

Taken together, the menopausal transition represents a period of heightened vulnerability for women, during which changes in hormonal signaling, immune tone, and brain metabolism converge. The decline in estrogen not only alters neuronal function and energy balance but also appears to shift the brain toward a more inflammatory state. These converging changes may help explain why women are disproportionately affected by AD, particularly in the years after menopause. Understanding how menopause interacts with genetic and immune risk factors will be critical for identifying the mechanisms underlying sex‐specific vulnerability to AD and for developing targeted interventions during this life stage.

6.3. Menpausal hormone replacement therapy (MHT)

A detailed evaluation of effects of MHT on AD risk and cognition is beyond the scope of this publication, and recent reviews are available. ¹⁹⁴ , ¹⁹⁵ , ¹⁹⁶ In summary, evidence for MHT's impact on cognition and AD risk is mixed. In women over age 65, use of MHT is associated with worse cognitive outcomes and increased risk for AD and dementia, particularly with prolonged use. ¹⁹⁷ , ¹⁹⁸ , ¹⁹⁹ MHT appears more beneficial when initiated before age 60 or within 5 years of menopause. ¹⁹⁹ , ²⁰⁰ Multiple studies report reduced risk of AD after MHT, ranging from 11% to 33%, depending on timing, formulation, and population studied. ²⁰¹ , ²⁰² , ²⁰³ Beneficial effects of MHT on indicators of disease in women with AD or at risk for AD include improvement in cerebral blood flow, cortical metabolism, and some measures of cognitive performance (Alzheimer's Disease Assessment Scale Cognitive subscale), but not others (Mini‐Mental State Examination, Clinical Dementia Rating). ²⁰⁴ , ²⁰⁵ , ²⁰⁶

Although the cognitive and metabolic effects of MHT have been studied extensively, albeit with conflicting findings, its immune effects remain understudied in the context of AD. Some peripheral immune changes have been reported, including increased B cell counts, enhanced T cell mitogen‐activated proliferation, and reduced natural killer cell–mediated cytotoxicity after MHT. ²⁰⁷ , ²⁰⁸ These changes are viewed as beneficial because they are opposite in direction to age‐related reductions in the same parameters. However, findings are inconsistent: one study reported that conjugated equine estrogens increased circulating CRP, ²⁰⁹ but another did not. ²¹⁰ Notably, MHT was associated with reduced hospitalization and mortality from COVID‐19 in a large study, suggesting possible systemic immunomodulatory effects. ²¹¹ As of yet, no studies have examined MHT's impact on brain immune activation using TSPO PET imaging in the context of AD.

7. CONCLUSIONS AND FUTURE DIRECTIONS

Sex differences in AD risk, progression, and clinical expression are increasingly recognized as fundamental, not incidental. Women are disproportionately affected by AD and show heightened tau pathology, stronger links between pathology and cognition, and a faster rate of decline for a given level of neuropathology. These disparities cannot be explained solely by longevity or diagnostic bias; rather, they reflect complex, interacting biological mechanisms. The evidence reviewed here supports a converging framework in which sex‐specific neuroendocrine function, immune signaling, and genetic and epigenetic regulation intersect to shape vulnerability to AD.

A central finding across multiple domains is that the neuroimmune environment is differentially regulated by sex steroid hormones. Estrogens and progesterone modulate glial cell function, inflammatory signaling, and neuroprotection, but their effects are neither universal nor uniformly protective. In microglia, estrogens suppress inflammatory cytokine release and promote cellular homeostasis, effects that may be diminished in the context of APOE ε4/ε4 genotype. Astrocytes respond dynamically to hormonal shifts, especially across reproductive transitions, influencing glutamate handling, calcium regulation, and injury response. These hormone–glia interactions are central to sex‐differentiated neuroinflammation, yet remain incompletely understood in the context of aging and AD.

Sex also interacts with genetic architecture in meaningful ways. APOE ε4 confers greater AD risk and tau‐related pathology in women, while ε2 may offer more protection in females than males, though findings are mixed. Beyond APOE, recent studies reveal sex‐specific associations at other loci related to amyloid processing, tangle burden, memory performance, and immune function. Importantly, many of these effects persist independently of circulating hormones, suggesting a role for epigenetic programming in sex‐biased gene expression, particularly in microglia. Female‐derived microglia maintain distinct transcriptional profiles even when transplanted into male brains, pointing to intrinsic, developmentally established regulatory mechanisms that influence disease trajectory.

Outside the brain, systemic immune processes also differ by sex. Women exhibit stronger associations between peripheral inflammatory markers and AD biomarkers, as well as broader neuroinflammatory gene expression across brain regions in aging. These sex‐specific immune signatures are further shaped by life stage. Pregnancy and menopause represent critical windows during which hormonal shifts may reprogram immune tone and neural resilience. Pregnancy may confer lasting immune modulation, while menopause marks a decline in estrogenic regulation of glia, coinciding with increased inflammation, altered metabolism, and enhanced vulnerability to neurodegeneration. To close the gap in our understanding of sex differences in AD, future research must embrace integrated, mechanistic, and sex‐aware approaches across molecular, cellular, systemic, and clinical domains. Based on the evidence reviewed, we propose the following research priorities:

7.1. Longitudinal, sex‐stratified biomarker studies across life stages

There is a pressing need for prospective studies that track hormonal, inflammatory, and neuropathological changes across critical life stages, particularly during menopause, pregnancy, and postpartum periods. Such studies should incorporate serial measures of glial activation, tau and Aβ burden, neuroimmune signaling, and cognitive function, with stratification by sex, APOE genotype, and hormonal status. To clarify the role of menopause in AD vulnerability, longitudinal studies should track hormonal, immune, and biomarker changes across the menopausal transition, while controlling for age and reproductive history.

7.2. Single‐cell, multi‐omic profiling of human glia by sex and genotype

Bulk tissue analysis masks the cell type–specific effects of sex, hormones, and genotype. Future work should prioritize single‐cell RNA sequencing, spatial transcriptomics, epigenomic profiling, and proteomics of microglia and astrocytes from male and female brains, ideally across life stages and disease stages. Human post mortem studies, iPSC‐derived glial models, and brain organoids should also be leveraged to dissect how sex steroid receptors, APOE variants, and inflammatory mediators shape glial phenotype in a sex‐dependent manner.

7.3. Mechanistic models incorporating sex, genotype, and hormonal milieu

Both preclinical and clinical studies must adopt intersectional frameworks that model the interactions among sex, genotype (e.g., APOE ε2 or ε4 carrier), and endocrine context (e.g., menopausal status, hormone therapy). Rodent models should include both sexes and manipulate hormonal exposures in combination with genetic risk alleles. Clinical trials should analyze sex–genotype interactions explicitly, avoiding the common practice of adjusting for sex as a covariate without testing interactive effects.

7.4. Investigation of epigenetic mechanisms driving sex‐specific vulnerability

Epigenetic programming likely underpins many of the observed sex differences in microglial function and AD risk. Studies should explore sex‐biased DNA methylation, histone modification landscapes, and non‐coding RNA expression (including sex‐linked microRNAs) in microglia and astrocytes. Importantly, these analyses must extend beyond animal models to include human tissue and patient‐derived cells, focusing on epigenomic features that persist independent of hormone exposure.

7.5. Development of sex‐specific therapeutic strategies

Precision medicine in AD will require sex‐stratified intervention development. Promising avenues include selective estrogen receptor modulators, glial‐targeted immunomodulators, and metabolic interventions timed to life stage. Trials of MHT must incorporate timing, formulation, and genotype as core design variables. For men, interventions may need to address age‐related testosterone decline or estrogen dysregulation. A “one‐size‐fits‐all” approach to neurodegenerative therapy is no longer tenable.

Together, these research priorities emphasize a need to move beyond superficial sex comparisons and toward mechanistically grounded, life stage–aware approaches to AD. By integrating sex, genotype, hormonal status, and immune context into the design and interpretation of both preclinical and clinical studies, the field can better understand the biological roots of disease heterogeneity. Only through such integrative, multidimensional strategies can we advance toward truly personalized prevention and treatment strategies that reflect the unique physiological trajectories of both women and men across the lifespan. This reconceptualization is essential not only for scientific accuracy but for equity in care, research, and therapeutic development in AD and related dementias.

CONFLICT OF INTEREST STATEMENT

Brittani R. Price is a full‐time employee of GE Healthcare. Timothy J. Hohman is a member of the scientific advisory board for Vivid Genomics and is the deputy editor for Alzheimer's & Dementia: Translational Research and Clinical Intervention, but had no involvement in review of this manuscript. There are no conflicts of interest to disclose for other authors. Author disclosures are available in the supporting information.

CONSENT STATEMENT

Previously published literature is reviewed herein. No new informed consent is necessary.

Supporting information

Supporting Information

Price BR, Walker KA, Eissman JM, et al. Sex differences and the role of estrogens in the immunological underpinnings of Alzheimer's disease. Alzheimer's Dement. 2025;11:e70139. 10.1002/trc2.70139