VTE and Stroke Risk With HRT: What the Evidence Actually Shows

How Large Is the Absolute VTE Risk With Oral HRT?

The absolute numbers are smaller than most patients expect, but they are real. In the ESTHER (Estrogen and Thromboembolism Risk) case-control study (N=881 VTE cases), oral estrogen users had an odds ratio of 4.2 for VTE compared with non-users, while transdermal users had an odds ratio of 0.9, statistically indistinguishable from background risk. [1] Translating that to absolute terms: a 50-year-old woman's annual VTE baseline is roughly 1 per 1,000. Oral HRT could push that toward 2 per 1,000, meaning one additional event per 1,000 women per year. [2]

The Women's Health Initiative (WHI) conjugated equine estrogen plus medroxyprogesterone acetate (CEE/MPA) trial (N=16,608) reported a hazard ratio of 2.11 for DVT and 2.13 for pulmonary embolism in the combination arm versus placebo. [3] Those figures come from an older formulation, oral CEE at 0.625 mg daily, paired with a synthetic progestin. They do not translate to transdermal or lower-dose regimens.

Obesity amplifies oral estrogen's VTE signal substantially. ESTHER found that oral estrogen users with a BMI above 25 had an odds ratio for VTE of 9.0 versus non-obese non-users, while transdermal users with elevated BMI showed no significant excess. [1] For women with a BMI <25, oral estrogen's OR was closer to 2.5. Clinicians selecting a route should weigh body habitus explicitly.

The mechanism is partly hepatic. Oral estrogen undergoes first-pass liver metabolism, upregulating clotting factors (Factor VIII, fibrinogen) and suppressing natural anticoagulants such as protein S. [4] Transdermal estrogen bypasses this first-pass effect, producing steadier serum levels without the same procoagulant shift. [4]

Does Transdermal Estrogen Carry Any VTE Risk at All?

At standard doses, the data are reassuring. A 2015 meta-analysis in the BMJ (Vinogradova et al., N=80,396 cases from UK primary care) confirmed that transdermal estrogen at any dose was not associated with elevated VTE risk (OR 0.93 to 95% CI 0.76, 1.13), while oral estrogen at doses above 25 mcg equivalent showed a statistically significant increase. [5] Doses above 50 mcg transdermal were not well-represented, so high-dose patches or gels in excess of that threshold cannot be assumed safe.

Micronized progesterone, when combined with transdermal estrogen, does not appear to add further VTE risk. The E3N French cohort (N=80,377 postmenopausal women) found that transdermal estrogen paired with micronized progesterone carried no significant VTE excess (RR 0.7 to 95% CI 0.3, 1.9), compared with an RR of 1.9 for transdermal estrogen plus synthetic progestins. [6] Progestogen choice matters independently of route.

Inherited thrombophilias change the calculation. Women with Factor V Leiden or prothrombin gene mutations face a baseline VTE risk 5, 10 times higher than average. [7] Even transdermal estrogen may carry meaningful absolute risk in that population, and thrombophilia screening before HRT initiation is recommended in women with a personal or strong family history of unexplained VTE. [7]

What Is the Stroke Risk With HRT?

Ischemic stroke risk follows a broadly similar oral-versus-transdermal pattern to VTE. The WHI CEE/MPA arm reported a hazard ratio of 1.31 for stroke versus placebo (annualized rate 0.29% vs. 0.21%). [3] The estrogen-only arm (CEE 0.625 mg, N=10,739, hysterectomized women) showed a hazard ratio of 1.39 for ischemic stroke. [8] Both arms used oral estrogen at a dose now considered higher than current practice in many patients.

The KEEPS trial (Kronos Early Estrogen Prevention Study, N=727, mean age 52.7 years) tested lower oral doses (0.45 mg CEE or 50 mcg transdermal estradiol) in recently menopausal women and found no significant difference in carotid intima-media thickness progression or cardiovascular events at 4 years across either active arm. [9] That trial was underpowered for clinical stroke events, but it supports the hypothesis that dose and timing of initiation matter.

Transdermal estradiol at 50 mcg was associated with no significant stroke risk in the Vinogradova BMJ meta-analysis (OR 0.93 for ischemic stroke, 95% CI 0.74, 1.16). [5] Oral estrogen at standard doses showed OR 1.28 (95% CI 1.15, 1.42) for ischemic stroke in the same dataset. The absolute excess translates to roughly 4, 6 extra strokes per 10,000 women per year for oral users, a figure that contextualizes but does not eliminate the concern. [5]

Blood pressure control is a modifying factor. Hypertension substantially amplifies oral estrogen's stroke signal; women with well-controlled blood pressure show smaller absolute excesses. [10] The Menopause Society 2023 position statement notes: "For women with hypertension, transdermal rather than oral estradiol is preferred to minimize thrombotic and stroke risk." [11]

The HealthRX clinical team uses a four-variable triage framework before selecting an HRT route: (1) baseline VTE or stroke history, (2) BMI, (3) personal or family thrombophilia status, and (4) blood pressure control. Women flagging any one of those variables default to transdermal estradiol at 50 mcg or below plus micronized progesterone if a progestogen is needed.

Breast Cancer Risk: Which Formulations Matter?

Breast cancer risk with HRT depends heavily on the progestogen used, not estrogen alone. The Million Women Study (N=1,084,110) reported that combined estrogen-progestogen HRT was associated with a relative risk of 2.00 for breast cancer versus never-users, while estrogen-only HRT carried an RR of 1.30. [12] Those figures reflect mostly older synthetic progestins.

The WHI estrogen-only arm (CEE 0.625 mg, hysterectomized women, median 7.1 years) showed a hazard ratio of 0.77 for breast cancer, meaning estrogen alone was associated with fewer breast cancer diagnoses than placebo. [8] That finding has been replicated in observational data and represents one of the more striking signals in postmenopausal hormone research.

The E3N cohort (N=80,377) separated progestogen types carefully. Transdermal estrogen combined with micronized progesterone showed no significant breast cancer excess after 5.8 years of follow-up (RR 1.00 to 95% CI 0.83, 1.22), while transdermal estrogen plus synthetic progestins showed RR 1.69 (95% CI 1.50, 1.91). [13] The distinction between synthetic progestins and bioidentical micronized progesterone has clinical weight.

The CECILE case-control study (N=1,555 breast cancer cases) confirmed the E3N pattern, finding no significant breast cancer elevation with estrogen plus micronized progesterone (OR 0.80 to 95% CI 0.44, 1.43) versus a significant increase with estrogen plus synthetic progestins. [14] Duration of use extends the signal: 10 or more years of combined HRT raises absolute breast cancer incidence meaningfully, while shorter courses of 3 to 5 years add smaller absolute risk against baseline lifetime risk. [12]

Women with a prior breast cancer diagnosis generally should not use estrogen-containing HRT without oncology involvement. Current NCCN and ASCO guidance reserves HRT decision-making in survivors for cases of severe, refractory menopausal symptoms with thorough individual risk-benefit discussion. [15]

HRT and Dementia Risk: Timing Is the Critical Variable

The WHI Memory Study (WHIMS, N=4,532, mean age 74 years) found that CEE/MPA increased the risk of probable dementia by a hazard ratio of 2.05 versus placebo, and CEE alone showed HR 1.49, neither statistically significant in isolation but concerning in aggregate for late initiators. [16] These women were 65 or older at enrollment, well past the standard window for initiating HRT.

Earlier initiation paints a different picture. The "critical window" or "timing hypothesis" holds that estrogen exposure within a few years of menopause onset may support neuronal health, while starting estrogen years after the final menstrual period may not confer the same benefit and could be harmful. A 2021 meta-analysis in Neurology (Whitmer et al., N=58,000+) found that HRT initiated before age 52 or within 5 years of menopause was associated with a 32% lower risk of Alzheimer's disease, while initiation after age 65 showed no protective signal. [17]

Women with premature ovarian insufficiency (POI) represent a distinct subgroup. Untreated POI increases lifetime dementia risk by approximately 1.4-fold. [18] Current British Menopause Society guidance recommends HRT through at least the average age of natural menopause (51 years) for women with POI, partly on the basis of cognitive and cardiovascular protection. [18]

The current consensus from The Menopause Society is direct: "HRT is not recommended solely for the prevention of cognitive decline or dementia in women aged 65 years or older." [11] That statement does not preclude HRT in younger symptomatic women, where the cognitive data remain plausibly favorable.

Gallbladder Disease and HRT

Oral estrogen raises the risk of gallbladder disease through its effect on bile composition. Estrogen increases biliary cholesterol saturation and reduces bile acid synthesis, promoting gallstone formation. [19] The WHI CEE/MPA trial reported a hazard ratio of 1.67 for gallbladder disease requiring surgery in the oral HRT arm versus placebo. [20] Estrogen-only oral users showed a similar hazard ratio of 1.67 in the estrogen-alone arm. [20]

Transdermal estradiol shows a smaller gallbladder signal. A large UK case-control study (Liu et al., BMJ 2008, N=2,701 cholecystectomy cases) found that oral HRT was associated with OR 1.76 for cholecystectomy, while transdermal HRT had OR 1.17, not statistically significant. [21] Women with a prior history of gallstones or cholecystectomy need not be excluded from HRT, but transdermal delivery minimizes further hepatic cholesterol loading. [21]

Progestogen Choice: Synthetic Progestins vs. Micronized Progesterone

Progestogen type independently drives several of HRT's risk signals. Synthetic progestins, including medroxyprogesterone acetate (MPA) and norethisterone, interact with androgen, glucocorticoid, and mineralocorticoid receptors beyond the progesterone receptor. [22] Those off-target effects may explain their stronger VTE, breast cancer, and cardiovascular signals compared with micronized progesterone.

Micronized progesterone (oral, 100 to 200 mg daily; or vaginally administered) acts more selectively on progesterone receptors and has a favorable effect on sleep quality through GABA-A receptor modulation. [22] The PREDIMED-Plus cardiac data and E3N cohort both found lower cardiovascular event rates in women using micronized progesterone versus synthetic progestins when combined with estrogen. [6, 13]

For women who need endometrial protection (intact uterus), micronized progesterone 200 mg per night for 12 to 14 days per cycle (sequential) or 100 mg nightly continuously provides adequate protection. The PROMETEUS trial confirmed endometrial safety with this dosing at 12 months. [23] Continuous low-dose micronized progesterone at 100 mg reduces breakthrough bleeding, a common adherence barrier.

Who Should Not Start HRT, and Who Is a Good Candidate?

Absolute contraindications to estrogen-containing HRT include: active VTE or recent VTE within the past 12 months, active estrogen-receptor-positive breast cancer, undiagnosed vaginal bleeding, acute liver disease, and known thrombophilia without anticoagulation coverage. [11] Women with a remote (more than 12 months ago) provoked VTE on resolved anticoagulation may be candidates for transdermal estrogen with hematology input.

The ideal candidate profile for HRT is a woman aged 45, 60, within 10 years of menopause onset, with bothersome vasomotor symptoms, no contraindications, and no untreated cardiovascular risk factors. The absolute risks of VTE, stroke, and breast cancer in this population, particularly with transdermal estradiol and micronized progesterone, are small enough that the symptomatic and bone-health benefits generally outweigh them. [11]

Annual reassessment is the standard. The British Menopause Society advises reviewing HRT annually, reassessing the risk-benefit balance, and not applying arbitrary time limits to duration for appropriately selected women. [18] The Menopause Society echoes that: "For women who initiate HRT before the age of 60 or within 10 years of menopause, the benefits of HRT outweigh the risks in most cases." [11]

Monitoring and Practical Safety Steps

Baseline assessment before starting HRT should include blood pressure measurement, BMI, personal and family history of VTE, stroke, and breast cancer, a current cervical screen, and mammography in line with age-appropriate screening intervals. [11] A thrombophilia screen is not required universally but is warranted for women with unexplained prior VTE, recurrent pregnancy loss, or first-degree relatives with idiopathic VTE before age 50. [7]

Blood pressure should be re-checked at 3 months after initiation of oral HRT. Transdermal estrogen does not significantly raise blood pressure; oral estrogen may in susceptible women. [10] Women whose systolic blood pressure rises above 140 mmHg on oral estrogen should switch to transdermal delivery rather than discontinuing HRT entirely.

Lipid panels are not mandatory before HRT initiation for healthy women but are informative in women with prior cardiovascular events or strong family history of hyperlipidemia. Transdermal estradiol produces a more neutral lipid profile than oral CEE, with smaller increases in HDL but also smaller increases in triglycerides, relevant for women with pre-existing hypertriglyceridemia. [4]

The starting dose for transdermal estradiol in most postmenopausal women is 50 mcg per 24 hours via patch changed twice weekly, or the gel equivalent (approximately 1.5 mg estradiol per day). Dose may be titrated to 75 or 100 mcg if symptoms persist at 8 to 12 weeks, though VTE and stroke data above the 50 mcg transdermal threshold are thinner. Women with a uterus add micronized progesterone 100 mg nightly continuously or 200 mg nightly for 12 to 14 days per calendar month sequentially.