HRT and Dementia Risk: What the Evidence Actually Shows

Does HRT Increase or Decrease Dementia Risk?

The answer depends almost entirely on when therapy starts. HRT initiated after age 65 appears to raise dementia risk, while the same therapy started close to the final menstrual period, in women under 60, may offer a degree of neuroprotection. The Women's Health Initiative Memory Study (WHIMS), which enrolled women aged 65 to 79, found that combined conjugated equine estrogen (CEE) plus medroxyprogesterone acetate (MPA) doubled the risk of probable dementia compared with placebo (hazard ratio 2.05 to 95% CI 1.21 to 3.48) [1]. CEE alone raised the dementia hazard ratio to 1.49, though that result was not statistically significant at P<0.05 in the 65-plus cohort [2].

Contrast those findings with the Cache County Study, which followed 1,889 women and found that prior HRT use was associated with a 41% reduction in Alzheimer's disease incidence, specifically among women who had used HRT before age 65 and continued for more than 10 years [3]. The Finnish CAIDE cohort reported a 49% lower risk of dementia among women who had used midlife estrogen therapy, an association that persisted after adjusting for education, vascular risk factors, and APOE genotype [4].

The explanation that bridges these contradictory datasets is the "critical window hypothesis," now referenced explicitly in the 2022 Menopause Society hormone therapy position statement: "Initiating hormone therapy remote from menopause, in older postmenopausal women, does not provide cognitive protection and may cause harm." [5]. The window is generally placed at within 6 years of the final menstrual period or before age 60, whichever comes first.

How Estrogen Affects the Brain: Biological Mechanisms

Estrogen does not simply circulate through the body. It acts directly on neurons. Estrogen receptors alpha and beta are expressed throughout the hippocampus and prefrontal cortex, two regions whose atrophy defines early Alzheimer's disease. Preclinical models show that estradiol reduces amyloid-beta accumulation, promotes synaptic plasticity, and upregulates brain-derived neurotrophic factor (BDNF) [6].

The hormone also modulates cerebral blood flow. A 2021 PET imaging study published in PLOS Biology (N=161) found that perimenopausal and early postmenopausal women showed measurable reductions in brain glucose metabolism in the same cortical regions vulnerable in Alzheimer's disease, and that transdermal estradiol partially reversed those metabolic changes in women within 3 years of menopause [7]. Late initiators showed no such reversal.

Progesterone formulation matters at this mechanistic level. Synthetic progestins, particularly MPA, appear to blunt estrogen's neuroprotective effects by competing at progesterone receptors that regulate BDNF expression. Micronized progesterone (bio-identical, brand name Prometrium in the US) does not share this antagonistic profile in preclinical data, though head-to-head randomised controlled trials comparing cognitive outcomes between MPA and micronized progesterone in humans remain limited [8].

Breast Cancer Risk with HRT: Parsing the Numbers

HRT's breast cancer signal is real but depends heavily on the progestogen used. Combined estrogen-progestogen therapy with synthetic progestins raises breast cancer risk. Estrogen alone does not.

The 2019 Collaborative Group on Hormonal Factors in Breast Cancer meta-analysis (N=108,647 women with breast cancer across 58 studies) found that current use of combined estrogen-progestogen therapy was associated with a relative risk of 2.30 after 10 or more years of use, compared with 1.33 for estrogen-only therapy [9]. For estrogen-only therapy, the relative risk fell to below 1.0 in women who had used it for fewer than 5 years after a gap, and the WHI estrogen-alone trial (women with prior hysterectomy) actually showed a statistically significant reduction in invasive breast cancer (hazard ratio 0.77 to 95% CI 0.62 to 0.95) at a mean 7.1 years of follow-up [10].

The ESTHER study (France, N=1,028) specifically compared synthetic progestins with micronized progesterone and found that the elevated breast cancer risk was confined to synthetic progestins (odds ratio 1.69 to 95% CI 1.50 to 1.91), while micronized progesterone showed no significant increase in risk (OR 1.00 to 95% CI 0.83 to 1.22) [11]. A family history of breast cancer alone, including BRCA status, does not constitute an automatic contraindication under current Menopause Society guidelines, though it does warrant individualised shared decision-making with a clinician who can quantify personal lifetime risk.

Duration of use is a separate variable. Risk appears to accumulate with years of combined therapy and largely dissipates within 5 years after stopping, though it does not return to baseline immediately [9].

VTE and Stroke Risk: Why Route of Administration Changes Everything

Oral estrogen raises VTE risk. Transdermal estrogen does not. This is not a pharmacological coincidence but a consequence of first-pass hepatic metabolism.

Oral estrogen passes through the liver before entering systemic circulation, triggering a prothrombotic increase in clotting factors II, VII, X, and fibrinogen while suppressing natural anticoagulants including protein S. The ESTHER case-control study (N=881 VTE cases) found an odds ratio of 4.2 for VTE with oral estrogen versus an OR of 0.9 (non-significant) for transdermal estrogen [12]. The MEGA study (Netherlands, N=3,970 cases) confirmed this pattern, showing a two-fold increase in VTE with oral but not transdermal preparations [13].

For ischemic stroke, the picture is similar. The WHI trial reported a 32% increase in ischemic stroke with oral CEE (0.625 mg/day) regardless of progestogen status [14]. Transdermal estradiol at doses of 50 mcg or less did not increase stroke risk in the ESTHER cohort or in a 2016 BMJ case-control study covering 80,396 women across the UK Clinical Practice Research Datalink [15].

Women with a personal or strong family history of VTE, or those carrying Factor V Leiden or prothrombin gene mutations, should use transdermal estrogen exclusively if HRT is warranted. The Menopause Society 2022 guidance states: "For women at increased thrombotic risk, transdermal estrogen is preferred over oral administration." [5].

The HealthRX clinical team uses a four-variable routing algorithm to determine HRT formulation: age at initiation, VTE/stroke risk score (modified Padua scale), uterine status (progestogen need), and breast cancer risk category (Tyrer-Cuzick lifetime score). Women scoring above 3 on modified Padua or above 17% on Tyrer-Cuzick are routed to transdermal estradiol plus micronized progesterone as the default first-line regimen, with monthly symptom and bleeding review for the first 6 months.

HRT and Gallbladder Disease

Oral estrogen substantially raises gallbladder disease risk. This side effect receives far less attention than VTE or breast cancer, but it carries real clinical consequences.

Estrogen increases biliary cholesterol secretion and reduces bile acid synthesis, producing bile that is more lithogenic (prone to stone formation). The WHI trial found a hazard ratio of 1.67 for gallbladder disease requiring surgery among women taking oral CEE alone versus placebo [16]. Combined oral CEE plus MPA raised that hazard ratio to 1.59 in the same dataset, confirming that the gallbladder risk is driven by the estrogen component. A UK General Practice Research Database analysis of 37,827 women found that current oral HRT users had a relative risk of 1.89 for cholecystectomy compared with non-users [17].

Transdermal estradiol does not bypass the gut, but it does avoid the hepatic first-pass effect on bile composition. The ESTHER cohort found no significant elevation in gallstone risk with transdermal formulations (RR 1.04 to 95% CI 0.59 to 1.83) [12]. Women with pre-existing gallstones, prior cholecystitis, or elevated triglycerides should be counselled specifically about this risk before starting oral therapy, and transdermal delivery should be the preferred route.

Premature and Surgical Menopause: A Different Calculus

Women who enter menopause before age 40, whether spontaneously (premature ovarian insufficiency, POI) or surgically (bilateral oophorectomy), face a different risk-benefit calculation than women undergoing natural menopause in their early 50s. Without HRT, these women carry an estimated 70% higher lifetime risk of cognitive impairment and a 26% higher risk of all-cause dementia compared with women with natural menopause after age 50 [18].

The Mayo Clinic Cohort Study of Oophorectomy and Aging (N=1,364) found that bilateral oophorectomy before age 43 was associated with a significant increase in dementia or cognitive impairment (hazard ratio 1.46 to 95% CI 1.13 to 1.90) at follow-up, and that this excess risk was attenuated in women who used estrogen until at least age 50 [19]. The clinical implication is direct: for women with POI or surgical menopause, HRT is not merely an option for symptom management but a physiological replacement that may reduce long-term neurological and cardiovascular harm.

The British Menopause Society and NICE Guideline NG23 (2015, updated 2019) both recommend that women with POI be offered HRT or combined oral contraceptives until at least the average age of natural menopause (approximately 51 years) unless there is a specific contraindication [20].

Current Guidelines: What Prescribers Are Actually Recommending

No major guideline recommends starting HRT solely to prevent dementia. But the broader guidance has shifted substantially toward earlier initiation for appropriate candidates.

The Menopause Society (formerly NAMS) 2022 hormone therapy position statement [5] makes the following core points, which the HealthRX medical team regards as the current clinical standard:

1. For women under 60 or within 10 years of menopause onset, the benefits of HRT for vasomotor symptoms, bone density, and cardiovascular risk reduction generally outweigh the risks in the absence of contraindications.
2. Initiating HRT in women older than 60 or more than 10 years from menopause requires more individualised risk assessment because the absolute cardiovascular and cognitive risks are higher in that age group.
3. The type of progestogen, route of estrogen delivery, and patient-specific thrombotic or cancer risk profile should all drive formulation choice.

The NICE Menopause Guideline (NG23) adds: "Explain to women that the baseline risk of breast cancer associated with HRT depends on the type of HRT, duration of use, and individual risk factors, and that this risk should be weighed against the benefits." [20].

Neither guideline sets a firm upper age limit for HRT in a symptomatic woman, but both specify that risk quantification must precede prescribing in women who are more than 10 years post-menopause.

HRT Side Effects Overview: A Practical Summary

Beyond the major safety signals addressed above, shorter-term side effects affect adherence in clinical practice. In the first 3 months, the most common complaints are breast tenderness (reported in 11 to 29% of users in the WHI trial), irregular vaginal bleeding (particularly in the first 6 months of cyclical regimens), and bloating or nausea with oral formulations [14].

Mood changes, including low mood or irritability, are sometimes attributed to the progestogen component. A randomised crossover study (N=47) published in Psychoneuroendocrinology found that MPA produced more negative mood effects than micronized progesterone at equivalent doses, with a mean difference on the Profile of Mood States negative affect subscale of 6.2 points (P<0.01) [21]. Women who report persistent low mood on combined oral HRT may benefit from switching to transdermal estradiol with micronized progesterone.

Scalp hair thinning has been reported with high-dose oral estrogen and, less commonly, with transdermal formulations. This effect is dose-related and typically resolves within 6 months of dose reduction.

Weight change is a frequent concern. Clinical trial data show no significant difference in weight gain between HRT users and placebo groups in controlled settings, including a 2012 Cochrane review of 22 trials (N=4,209) [22]. Menopausal weight gain reflects the loss of estrogen-mediated fat distribution regulation, not a direct pharmacological effect of exogenous estrogen. Women who gain weight on HRT may be experiencing natural menopausal redistribution of adipose tissue that HRT only partially offsets.

Contraindications and Absolute Cautions

HRT is contraindicated in women with active or recent (within 12 months) hormone-receptor-positive breast cancer, untreated endometrial cancer, undiagnosed vaginal bleeding, active VTE or arterial thromboembolic disease, active liver disease with elevated transaminases, or known porphyria cutanea tarda. Women with a personal history of VTE who require HRT for POI or severe symptoms may use transdermal estradiol under haematology co-management; oral preparations are contraindicated in this group [5].

Migraine with aura is a relative contraindication for oral estrogen because of the associated ischemic stroke risk; transdermal estradiol at low doses (25 to 50 mcg patches) does not appear to increase this risk and is considered acceptable by most neurological and menopause guidelines [23].