Does Estrogen Cause Blood Clots and Strokes? What the Actual Trial Data Shows
"I looked it up and it said hormone therapy causes blood clots and strokes." That sentence, or some version of it, comes up in almost every conversation about starting estrogen after menopause. It's not an unreasonable thing to worry about, and it's not made up out of nothing. But the way it usually gets explained skips the detail that matters most: which estrogen molecule the data is actually about, along with two others, how it gets into your body and when therapy starts.
All three of those have been tested directly, not just theorized about, and the actual data tells a more specific, more useful story than "estrogen is risky."
Not All Estrogen Is the Same Molecule
Most of what shaped the popular fear of "estrogen causes clots and strokes" traces back to the Women's Health Initiative, and the WHI used oral conjugated equine estrogens, sold as Premarin, a mixture of estrogen compounds derived from pregnant mares' urine. That's chemically and metabolically distinct from oral estradiol, the bioidentical form structurally identical to the estrogen your own body produced before menopause.
A study published in JAMA Internal Medicine tested this distinction directly. Researchers took advantage of a health system's switch from prescribing oral conjugated equine estrogens to oral estradiol in 2005 and compared cardiovascular outcomes between the two groups of postmenopausal women. Oral conjugated equine estrogen carried roughly double the risk of venous thromboembolism compared to oral estradiol, a statistically significant difference. Stroke risk did not differ significantly between the two oral forms. There was also a non-significant trend toward fewer heart attacks with oral estradiol. Among a subset of women with blood drawn, those on conjugated equine estrogen showed a stronger degree of activated protein C resistance, a laboratory marker of increased clotting tendency, than those on estradiol (Smith NL et al., JAMA Internal Medicine, 2014 — PMID: 24081194).
That's a meaningful distinction sitting inside the single word "oral." Conjugated equine estrogen and estradiol are not interchangeable at the level of hepatic and coagulation biology, and a warning built primarily on WHI-era data is, in a meaningful sense, a warning about conjugated equine estrogen specifically, not about oral estradiol as a category.
Why Route of Administration Also Matters
Oral estrogen, a pill swallowed and absorbed through the gut, travels straight to the liver before it reaches the rest of the body. That's called first-pass hepatic metabolism, and it's not a minor detail, it's the actual mechanism behind the clotting risk. When estrogen passes through the liver in that concentrated first pass, it stimulates the liver to increase production of several clotting factors and other proteins involved in coagulation, an effect that's well established in the pharmacology of oral estrogen generally, seen consistently across oral contraceptives and oral hormone therapy alike. That effect applies to any oral estrogen, regardless of which molecule is swallowed, but the magnitude isn't identical across formulations, which is the piece the section above fills in.
Transdermal estrogen, delivered through a patch, gel, or cream absorbed through the skin, enters the bloodstream directly and reaches the liver at normal circulating concentrations, the same way the estrogen your body produced before menopause did. It skips that concentrated first pass entirely.
A systematic review and meta-analysis of 15 studies quantified the difference route makes. Oral estrogen was associated with a 63% higher risk of a first venous thromboembolism, a blood clot, compared to transdermal estrogen, and a single case-controlled study included in the review also suggested a modest increase in stroke risk with oral use (Mohammed K et al., Journal of Clinical Endocrinology & Metabolism, 2015 — PMID: 26544651). The authors were direct about the evidence quality: this is observational data warranting low to moderate confidence, not a head-to-head randomized trial of oral versus transdermal route. It's also worth naming what this particular analysis can't tell us: the included studies spanned decades and didn't consistently break results out by which oral molecule was used, so this "oral" figure is a blend across formulations rather than an estradiol-specific number. Given the CEE-versus-estradiol gap shown above, that blended figure likely overstates the risk for a woman specifically taking oral estradiol.
What the Randomized Trials Actually Found
Two of the most carefully designed randomized trials in this area tested hormone therapy directly against placebo in women who started treatment close to menopause, rather than years later. They asked somewhat different questions, and it's worth being precise about which one measured what.
The Danish Osteoporosis Prevention Study randomized just over 1,000 recently postmenopausal women, average age 50, to hormone therapy or no treatment and followed them for up to 16 years. This trial did track actual clinical events, not just imaging markers. Women on hormone therapy had a significantly reduced risk of the combined outcome of death, heart failure, or heart attack, with no significant difference in stroke risk between groups, and no meaningful increase in blood clots (Schierbeck LL et al., BMJ, 2012 — PMID: 23048011).
Here's the detail worth sitting with rather than glossing over: the hormone therapy used in this trial was oral estradiol, not conjugated equine estrogen, and the distinction above is a real part of why that null clotting and stroke finding makes physiological sense rather than reading as a contradiction. This trial also enrolled young, recently menopausal, healthy women, the population with the lowest baseline clotting risk to begin with, and with roughly 500 women per group, it didn't have the statistical power to reliably detect a real but very modest difference in a rare event like stroke or VTE even if one existed at the smaller magnitude estradiol carries. Both things are true at once: oral estradiol likely still carries somewhat more clotting risk than transdermal estradiol, for the hepatic reasons described above, and it appears to carry meaningfully less than conjugated equine estrogen does, which is a large part of why a trial using oral estradiol in a low-risk population didn't turn up a signal. "No significant difference" here doesn't mean no risk exists, but it also isn't the coincidence it would be if oral estradiol behaved like conjugated equine estrogen.
It's also worth naming the other side of that first-pass liver effect, since it isn't only a downside. The same hepatic metabolism that increases clotting factor production also shifts cholesterol handling in a favorable direction, oral estradiol tends to lower LDL and raise HDL more than transdermal estradiol does, because that effect runs through the same liver pathway. That's a genuine two-sided tradeoff, not a reason to dismiss the clotting side of the equation, and it's a plausible part of why the DOPS trial, using oral estradiol, found a significantly reduced composite mortality, heart failure, and heart attack outcome. None of this makes oral estradiol risk-free or automatically the better choice for a given patient, but it does mean the honest comparison is oral estradiol versus transdermal estradiol on a specific set of tradeoffs, not "oral estrogen" as a single undifferentiated risk category.
The ELITE trial randomized postmenopausal women to estradiol or placebo and specifically tested what's called the timing hypothesis, whether starting hormone therapy soon after menopause produces different vascular effects than starting it a decade or more later. It's important to be clear about what this trial actually measured: its primary endpoint was carotid intima-media thickness, an ultrasound measurement of early artery-wall thickening, not clinical stroke or blood clot events directly. Women who started estradiol within six years of menopause showed slower progression of that arterial thickening compared to placebo, an effect that was not seen in women who started more than ten years after menopause (Hodis HN et al., New England Journal of Medicine, 2016 — PMID: 27028912). That's a meaningful signal about vascular health and timing, and it's a different kind of evidence than a trial counting actual stroke events, worth knowing the difference between the two rather than treating all "positive" trial results as answering the same question.
Together, these trials point to a more precise picture than a blanket risk statement: which specific estrogen molecule is being used meaningfully changes the clotting picture, route of administration changes it further within a given molecule, and timing after menopause meaningfully changes the vascular picture, and none of those three variables, nor the difference between a clinical-outcome trial and a surrogate-marker trial, gets mentioned in the version of this warning that circulates informally.
The question isn't whether hormone therapy carries any risk. Nearly every effective medication does. The question is which formulation, which route, and which starting point the data actually supports, and that's a more specific conversation than most people ever get to have.
A Familiar Pattern
A patient in her early fifties, two years past her final period, had done her own research before her first visit and arrived with a firm position: no hormone therapy, she'd read too many warnings about clots and stroke. Her symptoms, disrupted sleep, hot flashes several times a day, and a noticeable drop in her ability to concentrate at work, had been getting worse for months. Going through the actual mechanism together, why route matters, what the randomized trial data on early initiation actually showed, and what a transdermal option specifically would mean for her individual risk profile, changed the conversation from a blanket yes-or-no into a real decision. She started a low-dose transdermal estradiol, was monitored on the standard schedule, and her sleep and concentration measurably improved within two months.
Who Should Have This Conversation
This is relevant for anyone who has avoided hormone therapy specifically because of clotting or stroke concerns, particularly if menopause was within the last several years, since that's the window where the timing-hypothesis data is strongest. It's also worth revisiting if you were only ever offered oral estrogen and never had the route-of-administration distinction explained, since that's a different risk profile than a blanket "hormone therapy" warning implies. Anyone with a personal or family history of blood clots, stroke, or certain clotting disorders still needs an individualized risk assessment before starting any formulation, oral or transdermal, since these trial populations don't represent every individual risk profile.
What Monitoring and a Realistic Timeline Look Like
Route and dose get selected based on your individual risk factors and history, not defaulted to one option automatically. Baseline evaluation includes a personal and family clotting history before starting. Most patients notice symptom changes, sleep, hot flashes, mood, within four to eight weeks, with follow-up evaluation and periodic lab monitoring built into the plan for as long as therapy continues. This isn't a start-it-and-forget-it decision on either side, the route chosen or the ongoing monitoring.
If clotting or stroke concerns have kept you from having a real conversation about hormone therapy, that conversation should include the actual mechanism and the actual trial data, not just the headline version of the warning.
Has a health concern you read about online ever kept you from asking your doctor a more specific question?
A free consultation is a good place to start that conversation, no commitment required, just a chance to look at the full picture together.
— Dr. Darrell Wilcox | @wellnessdoc_4everyoung on Instagram
References
Mohammed K, Abu Dabrh AM, Benkhadra K, Al Nofal A, Carranza Leon BG, Prokop LJ, Montori VM, Faubion SS, Murad MH. Oral vs Transdermal Estrogen Therapy and Vascular Events: A Systematic Review and Meta-Analysis. Journal of Clinical Endocrinology & Metabolism. 2015;100(11):4012-4020 — PMID: 26544651 (observational evidence, low-to-moderate confidence per the authors).
Smith NL, Blondon M, Wiggins KL, Harrington LB, van Hylckama Vlieg A, Floyd JS, Hwang M, Bis JC, McKnight B, Rice KM, Lumley T, Rosendaal FR, Heckbert SR, Psaty BM. Lower Risk of Cardiovascular Events in Postmenopausal Women Taking Oral Estradiol Compared With Oral Conjugated Equine Estrogens. JAMA Internal Medicine. 2014;174(1):25-34 — PMID: 24081194.
Schierbeck LL, Rejnmark L, Tofteng CL, Stilgren L, Eiken P, Mosekilde L, Køber L, Jensen JB. Effect of hormone replacement therapy on cardiovascular events in recently postmenopausal women: randomised trial. BMJ. 2012;345:e6409 — PMID: 23048011.
Hodis HN et al. Vascular Effects of Early versus Late Postmenopausal Treatment with Estradiol. New England Journal of Medicine. 2016;374:1221-1231 — PMID: 27028912.
This content is for educational purposes only and does not constitute medical advice. Hormone therapy carries individual risks that vary by formulation, route of administration, dose, timing relative to menopause, and personal and family medical history, including clotting disorders. Any decision to start hormone therapy should be made individually with a qualified physician after a full evaluation. Individual results vary, and no outcome is guaranteed. Dr. Wilcox is licensed to practice in Texas, Arizona, Colorado, Oregon, and Florida.
